| Summary: | [GM45] black screen when laptop lid reopened: primary A assertion failure (expected on, current off) | ||
|---|---|---|---|
| Product: | DRI | Reporter: | Adric Blake <promarbler14> |
| Component: | DRM/Intel | Assignee: | Intel GFX Bugs mailing list <intel-gfx-bugs> |
| Status: | CLOSED FIXED | QA Contact: | Intel GFX Bugs mailing list <intel-gfx-bugs> |
| Severity: | major | ||
| Priority: | medium | CC: | intel-gfx-bugs, plantroon |
| Version: | DRI git | ||
| Hardware: | x86-64 (AMD64) | ||
| OS: | Linux (All) | ||
| Whiteboard: | Triaged | ||
| i915 platform: | GM45 | i915 features: | |
| Attachments: | |||
|
Description
Adric Blake
2018-08-15 02:26:52 UTC
Can you also verify if same issue is on latest drm-tip: https://cgit.freedesktop.org/drm-tip and send dmesg with drm.debug=0x1e log_buf_len=4M? Please add all logs plain text not packed up. I tested the drm-tip kernel on Aug 15, but I could not replicate the bug at hand.I bisected to what appeareds to be a merge into the 4.18(rc) drm-tip tree. Any commit during bisection that was based on a 4.17 kernel did not show the bug, though I can't determine whether the bug originated before (or because of) the 4.18 merge. Or perhaps Git is confusing me, which is also a possibility. But as an example, commit 05c72e77 ("drm/i915: Nuke the LVDS lid notifier") did not show the bug, even though it seems to be some ancestor of the buggy commit.
From what I can make out...
old buggy commit 294f96ae8aa5: Merge tag 'drm-misc-next-2018-07-18' of git://anongit.freedesktop.org/drm/drm-misc into drm-next
bisected-to working commit ef8e0ff97ae8: Merge tag 'drm-intel-next-2018-07-19' of git://anongit.freedesktop.org/drm/drm-intel into drm-next
That's the smallest change that I can confirm contains a fix and can determine at this point in time. I don't know when the bug got introduced.
No logs since the bug didn't appear, though I made them just in case.
The bug still exists in the 4.18.5 kernel, I forgot to add... Thanks Adric! Did you manage to reproduce the problem with latest drm-tip? If yes, please attach the logfiles that Jani asked for. commit 05c72e77ccda89ff624108b1b59a0fc43843f343 Author: Ville Syrjälä <ville.syrjala@linux.intel.com> Date: Tue Jul 17 20:42:14 2018 +0300 drm/i915: Nuke the LVDS lid notifier Going to assume that one fixed this. Please reopen if that isn't the case. Adric, can you confirm if this issue is resolved for you based on the above commit changes? Linux stable (4.18.6) doesn't work (bug is still there). Linux from torvalds git (as of yesterday) works, so Linux 4.19 should be fixed. Linux from drm-tip git (as of yesterday) also works. commit 05c72e77ccda89ff624108b1b59a0fc43843f343 Author: Ville Syrjälä <ville.syrjala@linux.intel.com> Date: Tue Jul 17 20:42:14 2018 +0300 drm/i915: Nuke the LVDS lid notifier appears to be applied to linux stable, so I don't think that commit was the one to fix the bug. I'll reopen until 4.18.y stable is fixed (or until 4.19 comes out). (In reply to Adric Blake from comment #7) > Linux stable (4.18.6) doesn't work (bug is still there). Because the fix isn't included there. Resolving back to fixed. Adric, can you please verify this with latest kernel? Adric, have you got any results from latest kernel? Created attachment 142661 [details]
drm.debug=0x1e dmesg; actions taken: boot, close lid, open lid, dpms force on, close lid, open lid, result=blank screen with cursor visible
I am getting the same behavior as described in the original report.
system info: Debian Stretch 9.6 on Thinkpad X200, GPU: GMA4500
xserver-xorg-video-intel version: 2:2.99.917+git20161206-1
affected kernel: 4.9.0-8-amd64
last working kernel: 4.9.0-7-amd64
kernel 4.18.0-0.bpo.1 from unstable is also affected
The fix is not in the 4.18 kernel as I have been made aware. It is in the 4.19 kernel. If you can get your hands on the 4.19 kernel and test whether that works, then it would merit creating a new bug report if the bug still exists. For the record, the problem has not happened again on my GM45 system with kernels 4.19 onward. I installed 4.19-0-trunk from unstable repository and the bug is gone. What are my chances of getting a fix backported to 4.9 stable releases? For now, I am sticking with the last working kernel from stable. The chances are very slim, since this is not a minor change in terms of code changes. If it could be backported so easily, the change would have at least made it into 4.18. As the drm/i915 codebase changes rather rapidly, a large version jump like that makes it very difficult for fixes for the latest kernel to cleanly apply to an old kernel like 4.9. > For the record, the problem has not happened again on my GM45 system with
> kernels 4.19 onward.
Closing based on this comment. Thanks for verifying!
|
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.